THE CHIEF inspector, it would seem, is seeking a change in role. Hard, scientifically provable facts are taking a back seat to a more generalised form of commentary.

Chris Woodhead has new concerns including pedagogical method - phonics and so forth - and even funding methodology, while his statistical pictures of the British school system have been less prominent.

This year’s annual report notably failed to include comparative analysis of the quality of teaching and learning - hitherto presented as a crucial area in the drive to raise standards.

Any retreat from facts could be prompted, in part, by growing doubts about the nature and reliability of the judgments of inspectors from the Office for Standards in Education. In the past these have always been presented as scientific evidence rather than as well-informed impressions.

Last week’s TES analysis of failing schools highlighted such doubts. It demonstrated that a substantial minority of failing schools - one in six - is actually progressing well in the two key areas of English and maths. One in four failing primaries appeared to be performing at a broadly average level, or better.

Attacks on OFSTED’s reliablity are nothing new. Professor Carol FitzGibbon’s value-added team at Durham University have unearthed school after school which, according to their results and social circumstances, are doing better than the inspectors say.

The point about The TES analysis is it used the benchmarks of OFSTED’s great rival, the Qualifications and Curriculum Authority. These set out broad categories of poverty (judged by free school meals) and the sort of results that schools within those categories should be aiming for.

According to these benchmarks, then, 18 per cent of failing schools are doing a better than average job in the two key areas of government policy, maths and English.

The apparent conflict between government agencies is the more surprising given that the QCA’s “benchmarks” have been issued to individual schools twice - once by the QCA, and once by OFSTED itself through its Performance and Assessment (PANDA) data.

OFSTED has long promised an investigation of the disparity between national curriculum test results and inspectors’ judgments and last week said that a report is imminent.

This will make a number of points. The first is that inspection looks at much more than test results, taking into account factors like the quality of management and pupils’ behaviour.

The report is also likely to claim that the small classes in primary schools make their test results unpredictable and therefore unreliable as indicators of overall performance. The match between secondary-school judgments and results is pretty good, said a spokesman last week.

Meanwhile, the QCA is gearing up for a significantly more sophisticated version of “value added” analysis than the benchmarks. No one expects the anomalies to go away.

POVERTY AND POOR RESULTS

The TES analysis also came up with alarming findings about the poverty in failing schools (as judged by eligibility for free school meals). Seven out of 10 had more than average levels while one in three suffered poverty at twice the national average, or worse.

If it is the case that “failing” schools are largely found among those with deprived intakes, it makes it harder for OFSTED to claim that there are many middle-class schools in dire need of a shake up. It also makes it difficult to claim that the roots of educational failure lie in poor teaching methods rather than in poor social circumstances. The detection of failure in a wider range of schools would lend greater credence to these claims.

CASE STUDY 1

Lewisham Bridge school, south London. 43 per cent of pupils eligible for free school meals. Results: 54 per cent of 11-year-olds reaching level 4 in English, 46 in maths. Lewisham Bridge is a typical inner-city primary combining a few white middle-class children with many in poverty. Two-thirds of the pupils have an ethnic minority background, some of them travellers. “It’s a massive injustice,” says Hilary Carnihan, head until last month. “The children were working so hard and did not deserve to be told they were failing.”

CASE STUDY 2

Archibald Primary School, Middlesbrough. 49 per cent of pupils on free school meals. Results: 58 for English and 66 for maths Both Archibald Primary and the new Middleborough education authority are challenging OFTED’s judgment. “This is a challenging area, but we have achieved good results,” says headteacher Pat Irving. “We were very surprised to be put under special measures.”